iDrive 6 + Keyless entry + BMW Connected Drive

[Bry5on]

It’s hard to put into words how much effort and willpower I’ve been putting into this. My boards were not designed appropriately. I found someone local to make iterative adjustments. I’ve been back to him twice. It’s just been so much of: try, determine what’s not working, schedule a repair time, drive out, drive back, test, determine what’s not working, repeat. You hit a wall where you realize it’s just not happening.

So I’m back to the drawing board, back looking for a designer, back to dealing with getting components. I’m really sad about it 😅

How messed up are the boards? I’ve done a few white wire fixes in my day and have a soldering setup here at home..

You have the schematics and the gerbers?

[YoitsTmac]

How messed up are the boards? I’ve done a few white wire fixes in my day and have a soldering setup here at home..

You have the schematics and the gerbers?

I have everything. The intent was to write firmware for these boards and then make mild changes. I’d have to tap into a tiny resistor to deliver 3.3V to the circuit. My designer forgot to connect it.

But after so many mild changes, I’d rather run a new batch with some updated changes: new cellular board and new gps module for better reception

[Bry5on]

I have everything. The intent was to write firmware for these boards and then make mild changes. I’d have to tap into a tiny resistor to deliver 3.3V to the circuit. My designer forgot to connect it.

But after so many mild changes, I’d rather run a new batch with some updated changes: new cellular board and new gps module for better reception

I mean we can at least get you to functional boards so you can start hardware in the loop testing of the software - then you're not stuck with a longer critical path. Just say the word!

[George Hill]

Definitely Bry5on, if you are just trying to prove the concept modify the board you have now. You might also find something else that needs to be changed for the future redesign.

I've got a couple DMEs repaired with a similar concept. Works just fine for your situation and easy peasy. IIRC this is solid core wire from a CAT5 bundle




And when we designed a PCB wrong, jumper to the rescue.

[YoitsTmac]

All right, everyone, I had a little bit of a freak out on Tuesday, or whenever I was having issues. I made what feels like a bazillion calls, and everything kind of magically fell into place. There are kind of two things at play: The first thing is I have these existing boards, which will now need their third iterative adjustment to them just to get them to work. The next thing is that, based on the firmware I've already tested, some things are really poor, including the GPS accuracy it's really good at saying you're at some place at the end of the cul-de-sac, but not really being able to have the accuracy strong enough to even say which building you'd be in or the left or the right side of a cul-de-sac.

So I did what any person would do in this situation, which is do a complete design overhaul. I've actually completely changed nearly everything on the board. I'm working with a local designer to implement a new cellular modem, a new GPS module, and even a new way to communicate with CAN bus. So there's very little that's left of the original design. I'm meeting with the new designer later today, or at least that's what he says. He said the same thing yesterday, but I understand we all have lives.

My biggest issue with my last design was how long it took to go from completed design to in my hands. While the PCB itself was made within a couple of days after submission, the assembly took 30 to 40 days, and then shipping took another 10 days. There's very few things that are more depressing than the thought of having to wait essentially two months for another prototype. So, thankfully, I found someone local who can do the assembly. Once the design is finalized, I have a good contact with the supplier, and I will begin the order there at the same time that I begin the manufacturing process of the PCB. And at the end of it, the PCB and the components required for the PCB should be arriving within a few days of each other. This should give me a new prototype within the next 30 days. Hopefully less. ..... But realistically, not.

[heinzboehmer]

And when we designed a PCB wrong, jumper to the rescue.

Even big companies do it! https://youtu.be/onxJqNzUyk0

My biggest issue with my last design was how long it took to go from completed design to in my hands. While the PCB itself was made within a couple of days after submission, the assembly took 30 to 40 days, and then shipping took another 10 days. There's very few things that are more depressing than the thought of having to wait essentially two months for another prototype.

Maybe I'm misreading, but are you subcontracting the board manufacturing and assembly to your designers? I've had pretty good experiences with the typical big fab houses (JLCPCB, Osh Park, etc.), both overseas and domestic. Time from clicking "buy" to having the boards in hand is usually 1-2 weeks.

If you want super duper fast prototyping, you can get some blank boards and find someone (or a maker space) with a desktop CNC to mill them out for you. You can buy special rivets to use in place of vias. Then assemble by hand and throw them on a hot plate. This method assumes you're designing two layer boards, of course.

[YoitsTmac]

Even big companies do it! https://youtu.be/onxJqNzUyk0



Maybe I'm misreading, but are you subcontracting the board manufacturing and assembly to your designers? I've had pretty good experiences with the typical big fab houses (JLCPCB, Osh Park, etc.), both overseas and domestic. Time from clicking "buy" to having the boards in hand is usually 1-2 weeks.

If you want super duper fast prototyping, you can get some blank boards and find someone (or a maker space) with a desktop CNC to mill them out for you. You can buy special rivets to use in place of vias. Then assemble by hand and throw them on a hot plate. This method assumes you're designing two layer boards, of course.

I already have a couple wires running around my board. It’s just so far off the spec that at some point you just have to admit it’s time to go back to the drawing board.

For this first board, I subcontracted the design designing to one person, and I handled the manufacturing subcontracting myself. For this next iteration, I handed a rough draft of my new design to my designer, we’re currently talking tonight getting the spec fully concrete.

From there, the current plan is to have the PCB done overseas and I have a component supplier in North America that I plan on using. So I can have the parts order being put together and in transit at the same time that the PCB is being assembled and in transit.

Then the idea is to have both the PCBs and the components arrive at roughly the same time, where my designer would then handle the assembly. He did an incredible job repairing my current boards, and I fully trust him with this step.

When I had the first design assembled overseas, I saw my board had been completed in a really timely manner, but the assembly took a very long time. Granted, my very first board design carried a component that was basically impossible to source. Which I definitely outlined here in this thread. So that exacerbated the issue, but even after the BOM was finalized, it took 3 to 4 weeks to complete assembly.

The board is actually somewhat complex. I don’t believe I have shared the entire spec yet because I have some extra bells and whistles that I want to be a surprise and first reveal at a software level. The last design had about 130 components.

[YoitsTmac]

Update because I've been thinking of one, short because I'm tired.

All my parts over five or six orders are officially all stateside and are being assembled tomorrow. I will be working on validting features and firmware tomorrow. So things should be moving really fast over the next two weeks. I'll make sure to include a lot of photos. I've prepped an entire developement pipeline to validate messages and communication. Actually, I wish there was more to share, but there's not much.

I also am now on V3 of my idrive bezel. I'm about 3mm short on the width, which is sadly the closes I've ever gotten. I think I'm only interfering with the dash inadvertantly in one place. I'm going to dremmel it and see if its a success. Fingers crossed!

[YoitsTmac]

Nothing but great news in this update. Today I drove to my assembler and picked up my one and only prototype. I have parts for more, but I wanted to get this revision done. There are a few small things here and there, but it is definitely enough to get testing. For the nerds out here, here's everything the board can do.
- Kbus communication
- CANBUS (up to two buses)
- GPS signal
- Cellular or Wifi for internet
- Bluetooth (settings and pairing)

I picked it up at 4 pm this afternoon, and I haven't been off my laptop since. I've been making small iterations to my firmware. I made a few libraries that need to be tested, and so far, all testing has gone incredibly smoothly. Going to make a couple of updates, such as the ability to turn off the modem for ultra-low-power states.

And against all odds, the good news doesn't stop there. I think I finally may be on a revision of the iDrive bezel that is worth sharing. I purchased a really good, high-quality scan of the dash and another high-quality scan of the dash trim. Admittedly, these textures are AI-generated based on pictures I gave AI of the dash and of my trim in my car. The rest of the interior is from a car model I extracted from a video game. But here are some renders of the iDrive bezel I plan on getting printed and shipped out with the next revision of my board.



I kind of realized that the iDrive knob is cool, but most people aren't going to want to move forward with it and give up the room in their center console. So instead of eight favorite buttons, the eighth button is actually a back button, and if you hold it down, it's a menu button. The right knob basically becomes the iDrive knob. That way, you can navigate around. You can click down, go back and return to the main menu. By no means is it a full replacement, but I think it's a pretty good job. And obviously, this decision wasn't motivated at all by my inability to make a good mount in the center console.

Anyone who wants to weigh in on the design is welcome to, but I personally am going to add the 3D camera view to my car with my iDrive. I'm debating if I should make the eighth button into two buttons above each other, but that would create asymmetry. But I need a button to toggle the cameras. Alternatively, what I was thinking about doing was getting a convertible set of buttons and using the convertible buttons to activate the camera. At least in my initial testing, I'm not entirely sure if these are broadcast over KBus or anything. I do have a set of buttons I can test against.

Expect some good updates over the next 5 to 15 days. Until then, enjoy these renders (again, I merged a real dash scan with a video game. I know its far from perfect but gets the idea).

[George Hill]

In the future if you need a scan I'd say ask here first, there are quite a few of us with scanners that I am sure would be happy to help so you don't have to buy something like that.

Is the LH rotary knob for volume? My gut reaction is to use that one as your IDrive "knob." My thought is its easier to reach than the RH and you will have the steering wheel controls to fine tune the volume. If you did need a quick large volume knob adjustment then you can reach for the RH knob. My thought is for my I almost never use the volume knob on the radio itself.

Your pics are a bit dark on my screen but I feel like your design is really good. This is nit picking, but only because you asked I just hope you can nail the correct LED backlight color to match the existing LEDs. And maybe its just the AI pics but backlighting around the knob is just a bit more "strong" than the rest and really draws your attention, if you could make it a little thinner or "softer" I think it would blend better.

But if not, its not the end of the world, you've done a great job on this project. I'm excited to see it progress!

[Obioban]

This looks gooooooood, Never thought I'd be desiring iDrive, but carplay on OE quality hardware has my wanter activated.

In the future if you need a scan I'd say ask here first, there are quite a few of us with scanners that I am sure would be happy to help so you don't have to buy something like that.

Is the LH rotary knob for volume? My gut reaction is to use that one as your IDrive "knob." My thought is its easier to reach than the RH and you will have the steering wheel controls to fine tune the volume. If you did need a quick large volume knob adjustment then you can reach for the RH knob. My thought is for my I almost never use the volume knob on the radio itself.

Your pics are a bit dark on my screen but I feel like your design is really good. This is nit picking, but only because you asked I just hope you can nail the correct LED backlight color to match the existing LEDs. And maybe its just the AI pics but backlighting around the knob is just a bit more "strong" than the rest and really draws your attention, if you could make it a little thinner or "softer" I think it would blend better.

But if not, its not the end of the world, you've done a great job on this project. I'm excited to see it progress!

I've been effectively using his setup for a while (right knob is carplay control, left knob is volume) with my xtrons, and it feels completely natural coming from any e46 head unit. I think it would cause a lot of required muscle memory relearning to do it the other way. Plus, 99% of the time you'll use it as a touch screen anyway, so having the volume knob closer at hand probably makes sense.

.. that said, he could probably make a configurable option.

[George Hill]

.. that said, he could probably make a configurable option.

You and I are tall so its likely less of a reach for us, but a shorter armed driver might be less comfortable making the reach was my thinking that I should have articulated better.

And agreed, so much hate for iDrive, but I love it. I'd never pick a Radio over iDrive, this is a killer project. 😎

[Bry5on]

Nothing but great news in this update. Today I drove to my assembler and picked up my one and only prototype. I have parts for more, but I wanted to get this revision done. There are a few small things here and there, but it is definitely enough to get testing. For the nerds out here, here's everything the board can do.
- Kbus communication
- CANBUS (up to two buses)
- GPS signal
- Cellular or Wifi for internet
- Bluetooth (settings and pairing)

I picked it up at 4 pm this afternoon, and I haven't been off my laptop since. I've been making small iterations to my firmware. I made a few libraries that need to be tested, and so far, all testing has gone incredibly smoothly. Going to make a couple of updates, such as the ability to turn off the modem for ultra-low-power states.

And against all odds, the good news doesn't stop there. I think I finally may be on a revision of the iDrive bezel that is worth sharing. I purchased a really good, high-quality scan of the dash and another high-quality scan of the dash trim. Admittedly, these textures are AI-generated based on pictures I gave AI of the dash and of my trim in my car. The rest of the interior is from a car model I extracted from a video game. But here are some renders of the iDrive bezel I plan on getting printed and shipped out with the next revision of my board.

I kind of realized that the iDrive knob is cool, but most people aren't going to want to move forward with it and give up the room in their center console. So instead of eight favorite buttons, the eighth button is actually a back button, and if you hold it down, it's a menu button. The right knob basically becomes the iDrive knob. That way, you can navigate around. You can click down, go back and return to the main menu. By no means is it a full replacement, but I think it's a pretty good job. And obviously, this decision wasn't motivated at all by my inability to make a good mount in the center console.

Anyone who wants to weigh in on the design is welcome to, but I personally am going to add the 3D camera view to my car with my iDrive. I'm debating if I should make the eighth button into two buttons above each other, but that would create asymmetry. But I need a button to toggle the cameras. Alternatively, what I was thinking about doing was getting a convertible set of buttons and using the convertible buttons to activate the camera. At least in my initial testing, I'm not entirely sure if these are broadcast over KBus or anything. I do have a set of buttons I can test against.

Expect some good updates over the next 5 to 15 days. Until then, enjoy these renders (again, I merged a real dash scan with a video game. I know its far from perfect but gets the idea).

The center console buttons are not broadcast over the k-bus - I retrofitted an e39 PDC button in to my console (I have front and rear PDC) years ago and reverse engineered the way the buttons work. There are definitely ways to hijack a button and pass its signal out though

Reverse signal is broadcast over k-bus. That, combined with CAN wheel speeds would get you auto-trigger on reverse, plus a software counter of distance driven and/or vehicle speed to automatically exit. I think that’s how the factory does it anyway.

Don’t stack the buttons! if needed, you can always hack your button 7 on the display to become the 3D camera view. If you really wanted to get fancy, add a switch/jumoer/solder pad to your button PCBA to bring that button out as an alternative function.

Glad you’re up and moving! I’m still motivated to find a way to fit the control unit behind the dash

[YoitsTmac]

In the future if you need a scan I'd say ask here first, there are quite a few of us with scanners that I am sure would be happy to help so you don't have to buy something like that.

Is the LH rotary knob for volume? My gut reaction is to use that one as your IDrive "knob." My thought is its easier to reach than the RH and you will have the steering wheel controls to fine tune the volume. If you did need a quick large volume knob adjustment then you can reach for the RH knob. My thought is for my I almost never use the volume knob on the radio itself.

Your pics are a bit dark on my screen but I feel like your design is really good. This is nit picking, but only because you asked I just hope you can nail the correct LED backlight color to match the existing LEDs. And maybe its just the AI pics but backlighting around the knob is just a bit more "strong" than the rest and really draws your attention, if you could make it a little thinner or "softer" I think it would blend better.

But if not, its not the end of the world, you've done a great job on this project. I'm excited to see it progress!

Thank you for the kind words! While I agree that it's a bit of a reach, I still think this is the right design architecture. I can make it customizable in software, but let's not forget that one of iDrive's greatest features here is that it's a touchscreen (which makes a lot of sense in our interior), and also that there are seven favorite buttons. Once you set all your favorite places to go, you shouldn't be needing the iDrive knob much at all. And of course, if you still want to have one in place of your cup holders, you can still do that as well.

.. that said, he could probably make a configurable option.

Yeah, the only difficulty is that the label on the buttons is permanent, so I'd have to think about that and probably bounce around a few ideas here. I definitely want to live with it for a bit.

The center console buttons are not broadcast over the k-bus - I retrofitted an e39 PDC button in to my console (I have front and rear PDC) years ago and reverse engineered the way the buttons work. There are definitely ways to hijack a button and pass its signal out though

Reverse signal is broadcast over k-bus. That, combined with CAN wheel speeds would get you auto-trigger on reverse, plus a software counter of distance driven and/or vehicle speed to automatically exit. I think that’s how the factory does it anyway.

Don’t stack the buttons! if needed, you can always hack your button 7 on the display to become the 3D camera view. If you really wanted to get fancy, add a switch/jumoer/solder pad to your button PCBA to bring that button out as an alternative function.

Glad you’re up and moving! I’m still motivated to find a way to fit the control unit behind the dash

So, how does the center console button for the convertible roof happen? It's got to be over CAN or k-bus, surely. Or is it seriously an analog signal? The button rows that you're seeing are physical buttons, and they are not software. I guess I could just have six favorite buttons for myself and have a camera button. Or have it labeled up to 7, but if you have a camera module hooked up, then the seventh button is your camera button. That might be the easiest way to go about it.

[Bry5on]

Thank you for the kind words! While I agree that it's a bit of a reach, I still think this is the right design architecture. I can make it customizable in software, but let's not forget that one of iDrive's greatest features here is that it's a touchscreen (which makes a lot of sense in our interior), and also that there are seven favorite buttons. Once you set all your favorite places to go, you shouldn't be needing the iDrive knob much at all. And of course, if you still want to have one in place of your cup holders, you can still do that as well.


Yeah, the only difficulty is that the label on the buttons is permanent, so I'd have to think about that and probably bounce around a few ideas here. I definitely want to live with it for a bit.




So, how does the center console button for the convertible roof happen? It's got to be over CAN or k-bus, surely. Or is it seriously an analog signal? The button rows that you're seeing are physical buttons, and they are not software. I guess I could just have six favorite buttons for myself and have a camera button. Or have it labeled up to 7, but if you have a camera module hooked up, then the seventh button is your camera button. That might be the easiest way to go about it.

The convertible PCBA is actually different if memory serves. I haven’t probed it but a quick WDS check would answer that. I wouldn’t bet against analog, it’s only two wires (three/four if there’s LED feedback).

And yes, the latter is what I meant. I’m assuming you’ll have a narrow, long PCBA that has the buttons and volume/iDrive trim pots on it. You can either have them software defined, or hardware defined. Hardware defined seems easier to me but I don’t know what button architecture you have in mind

For developing the bezel and buttons, investing in an SLA printer wouldn’t be a terrible idea. I expect there will be a lot of iteration there.